Okay so I really hate Apple and their tech is overpriced etc

Okay so I really hate Apple and their tech is overpriced etc.
But I can't deny this language is basically the future.
Firstly it Is still shit because it doesn't have stable ABI on Linux, Windows but this is confirmed to change in version 5.1.
After this it will basically dominate the industry.
> open source
> cross platform(Linux, Windows, macOS, Android etc.)
> syntax isn't shit
> literally llvm language which with proper optimisation can be even faster than C
> because of the speed google already choose it as the main tensor flow language which is supposed to replace C and Python
> The creator is literally Clang creator and thinks Swift is going to be most popular language in future
> Also the creator works in google so basically big G will support it everywhere(confirmed fuchsia native support)
> Already works on Apple platforms
> IBM and Amazon thinks it's future of web apps and invest in them heavily(check Kitura, Vapor, Perfect)
> google already made it's own fork for swift if Apple does something weird
> it can use all of python libraries
> the biggest future competitor was Rust(great language btw.) but nobody of the big companies chose it so it's basically over.(also the syntax is nearly the same)

Attached: PatrickBateman.jpg (770x578, 20K)

Other urls found in this thread:

stackoverflow.com/questions/24101718/swift-beta-performance-sorting-arrays
github.com/tensorflow/swift/blob/master/docs/WhySwiftForTensorFlow.md
benchmarksgame-team.pages.debian.net/benchmarksgame/faster/swift-rust.html
forge.rust-lang.org/platform-support.html
media.ccc.de/v/35c3-9670-safe_and_secure_drivers_in_high-level_languages
news.ycombinator.com/item?id=19519164
cs.utexas.edu/users/EWD/transcriptions/EWD08xx/EWD831.html
twitter.com/NSFWRedditVideo

>the creator this, the creator that
Listen, Swift has some good things going for it, but its creator is most definitely not one of them.

POO IN LOO

Attached: 1539026788025.jpg (2016x1210, 278K)

Lol worst libs support than Ocaml/Haskell or scheme.

Java vertx massive outperformance Swift non-libs web frameworks.

>literally llvm language which with proper optimisation can be even faster than C
C compiled with clang, not gcc.

I think you have never written a big C/C++ project if you don't know why it's removed(it's also removed most of new languages).
When you are going fast through someone else code you will sooner or later mistake pre-inc with post-inc because they look nearly the same and good programmers do their job fast.

Hard to tell but yeah gcc C should be faster.(depends on use case).
C++ will be always slower though(because of lack of CIL)

Attached: 1555668792716.jpg (1190x711, 48K)

this was supposed to be the answer to

There really is no need for these operations besides inlining inc and dec operations before compilers were smart enough to optimize them. Which is something that hasn’t been necessary for a long time. It’s a vestige of an old age, like base 0 indexing.

OH NO NO NO NO

Attached: 1530127247651.png (760x547, 42K)

Original question here:
stackoverflow.com/questions/24101718/swift-beta-performance-sorting-arrays
You can see for yourself it's hugely improved since the pre-1.0 days

>swift
>dominate cross platform industry

Attached: 1530896001156.jpg (628x628, 83K)

> not realising that market is dominated by chrome which is a fork of webkit.
The same will happen with swift(already is happening).

Attached: 1554060603102.gif (487x560, 898K)

WebKit is sepples tho
Swift only exists to replace ObjectiveC

>Webkit a fork of khtml

I mean in all honesty the difference between `x++` and `++x` is dumb enough I feel it shouldn't be used, I am sick and tired of people causing bugs by typing `myArr[x++]` to look smart, you saved on a whole new line, big fucking deal. Also compilers nowadays are smart enough to tell when you don't need the discarded value so yeah, case closed.

Nah Apple uses Swift to replace Obj-c.
But it's not created specifically for that(the language is too powerful for simple apps).
That's why tensorflow, IBM are so interested in it.

Yeah so I just tested it(Swift 5.0).
And the times are pretty great.
And yeah in original question it was beta of Swift 1.0. So bravo retarded user for posting that screenshot.

Attached: Screenshot 2019-04-22 at 02.57.05.png (2256x808, 684K)

Tensorflow is interested because Google wants to run ML on iPhones. IBM, ... is IBM. There is really no purpose for Swift beyond corralling developers into an environment managed by corporate titans. The language offers nothing you can't get elsewhere, and the ecosystem is undeveloped.

Tensorflow is mostly interested in optimization.
And there is a lot of reasons.
Here are the official ones.
github.com/tensorflow/swift/blob/master/docs/WhySwiftForTensorFlow.md

Attached: 57226346_10161622127770483_4075134377116303360_n.png (960x960, 763K)

It doesn't add up. I understand the problems with using python for scientific applications, those are well known. But their reasons for choosing Swift aren't convincing.
Their long list explaining why they didn't pick language X reads as a laundry list of nitpicks they came up with after the fact. A technical question mark is the lack of static functional languages, which would be perfect for the static analysis they require, but were seemingly not even considered. A community question mark is who the hell uses Swift besides Apple developers? They even mention that scientists don't use Swift, and its a pretty big ask to expect them to use a new ecosystem JUST to use Tensorflow (which most of them actively avoid for a variety of reasons to begin with).

Why would they need to use new ecosystem?
Swift is working on all platforms and with LSP there will be a lot of ide's working with it.
Google also gave people swift support in google collab.

Write in Lua and let's see how many useless lines you end up with. Basic shit like i = i + 1 doesn't need it's own line and you're brain damaged if you think it should.

Ecosystem means the set of tooling and packaging which support development in a language and platform. Going from Python to Swift means throwing away a lot of that and it's associated knowledge for something new for questionable reasons.

The reason Tensorflow is doing this is because they are losing to PyTorch in the deep learning community as a whole. Academics and ML experts just don't like Tensorflow. So they need a compelling reason for people to use their framework, and support for mobile development is a huge one.

>useless lines
If you're not using an OS that's less that 100k lines of code then you have no authority on teaching about bloat

literally nobody with a clue uses integer loop counters in the current year when the language itself has an iteration protocol

Python is now compatibile with Swift though.
You just create python object in swift code and can use all of python libraries. (rewriting them natively would still be useful though).
So they aren't really using anything.

Swift starts a python interpreter to run your python code. This is something you can achieve (albeit with some work) in any language. It's not magic.
You still need to write Swift for there to be any reason for using Swift to begin with. Which means now your managing two stacks of technology until you move over fully. That's a big ask from people who are more interested in experimentation and don't come from a software engineering background.

Funny you should mention that...

It's just an example variable. I should have made it "x".

>uses manchild fruit toys
>calls someone else retarded

goddamn macos is so fucking good i wish it wasnt on such shit laptops

> Call macbook users retards
> Most scientist use them

Attached: 1498105506505.jpg (640x360, 16K)

>most scientists are retards

Nothing Apple championed ever left the walled garden.

Sounds pretty good, but I think C# .NET Core would give it a run for its money.

>gcc C should be faster.
It isn't a lot of the time. LLVM is pretty good.

> Formatting text like (((this)))
> Not knowing that they only use MacBooks because macOS isn't getting in their way and/or they have programs that require it

> cross platform(Linux, Windows, macOS, Android etc.)

nobody use swift outside ishit

> literally llvm language which with proper optimisation can be even faster than C

like every new languages out there.

iTODDLERS BTFO

Attached: 1532486038629.gif (500x491, 376K)

>you can use swift to use python libraries which are just c++ bindings for python
>and that's a good thing

Everyday we stray further from God's light.

technology illiterates

RUST FAGS BTFO

Attached: video-game.jpg (3000x1997, 1M)

Does it run on OpenBSD yet? No? Okay, tell me when I can give a shit.
Same goes for Rust, actually.

JS, C++, Haskell, Bash
these are all of the languages you need

Attached: 1555708358590.png (359x414, 150K)

Not yet user, not yet. benchmarksgame-team.pages.debian.net/benchmarksgame/faster/swift-rust.html

forge.rust-lang.org/platform-support.html

>Tier 3 platforms are those which the Rust codebase has support for, but which are not built or tested automatically, and may not work. Official builds are not available.
Which is where OpenBSD is in.
It doesn't work.
Try reading what I said instead of mindlessly linking to that which I already know.

It is not *automatically tested*, that does not mean it does not work, unless you've tried and actually had an issue (in which case do report it, not many people will since the *BSD's are extremely niche).

>unless you've tried and actually had an issue (in which case do report it, not many people will since the *BSD's are extremely niche).
Yes, I have tried and actually had an issue. Dozens of them. At some point effort > payoff.

Rust is overrated.
What's the point in spending so much effort ensuring memory and thread safety, to the point of it being a bother, and then spend little effort on helping people write correct code? At least Ada figured out what its niche is and how to accommodate it with Spark. Rust will never take off because the community is obsessed with fixing Cs beginner pitfalls instead filling a real need.

Attached: D4KLqk2U4AALQ7z.jpg (1536x2048, 359K)

>what's the point in spending so much effort
You don't, unless you're a beginner to the language in which case yes the learning curve is very steep. I assume you're referring to the whole "wrestling with the borrow checker" garbage which is a myth spread around by angry newcomers who try to use it like C.

I'm talking about the developer effort that is put into these features and the meme safety cult the sprung up around it.

I don't see a problem with putting a lot of effort into said features as long as in the end they'll have a good product. It's definitely holding it back though in terms of development time.

The safety is pointless anyway, you can’t do anything fun with Rust unless you use unsafe blocks

t. has literally never used rust

meanwhile, in non-bizarro/n-gate world:
>open source, but so is a lot of stuff Oracle or SAP do
>cross platform, but it doesn't matter since it's still basically worthless without reference counting fagOS libraries
>syntax is good, well, at least it isn't indentation based and can be parsed
>bound to LLVMs mediocre compile times, while problems like arise and its memory management model makes sure it will never be fit for high-throughput scenarios
>some webshit advertisement company without creditability in programming languages or pretty much anything at all beside botnets figures out after 3 decades of hard research that C and Python are pure garbage and decide to replace it with some sub-mediocre middle ground, as they are used to
>some tool that created a compiler (they come dime a dozen) that created said epitome of mediocracy thinks its the future
>named pleb works for named big botnet company
>it happens to work for the platform for which it was deliberately crippled to make it work
>some obsolete, clueless IT investor and a logistic company that lot of retards fall for have some field research projects nobody heard of involving said technology
>botnet company knows well about foul play, therefore tries to avert foul play from "business partner"
>it can use libraries from a programming language that can't even check for spelling mistakes in identifiers, therefore perfectly providing the foundation for those fruity 10 liners professional devs are known to use in their companies pillar of succ
>the biggest competitor isn't Rust (an ok-tier language besides the hindsight bonus) which actually a lot of big companies use but other mediocre RAD languages that tear it a new one
There are some good ideas in Swift, OP, but other than that it is completely like Vala, or even Genie (Vala with indentation syntax) and if it wasn't for massive shilling from Apples (and this goes for most corporate languages) it would have shared its fate long ago.

fix:
>while problems like
thought the screenshot showed the benchmark where PYthon wins

Maybe. I am pretty sure C# is already a winner in game engines so the language has good future for sure.
Google, IBM, Amazon. Literally the biggest companies in the world.
After Swift 5.0 compile times are nice and it memory managment model was never a problem it's probably swifts best feature.
this was already proved to be false and was present in beta of swift 1.0 which had not stability at all. You even had user who proved it in this thread.
Vala was never more than a gimmick not adopted by any big players. Swift is adopted by big players for future of their big projects.

>Google, IBM, Amazon. Literally the biggest companies in the world.
Imagine believing this.
Now imagine considering this a good reason even if it was true (it’s not).

Attached: BE340F91-6F38-41FE-91D7-AFB3AFA44C37.png (888x720, 308K)

Attached: paul-hudson-twostraws-feeling-a-little-cold-just-put-these-46774358.png (500x486, 32K)

Nobody said it's a good reason. But language needs support of some big companies today to not be failure. It's no longer the eighties.
Creating a never ending reference cycle. Bravo. It's like being surprised that recursion which never reaches base case isn't ending.

Reminder that Swift 5.0 is first stable release and before it there were shitty things. All of these don't exist anymore.
This code simply won't run.

Attached: Screenshot 2019-04-22 at 13.06.24.png (1354x138, 29K)

>But language needs support of some big companies today to not be failure. It's no longer the eighties.
So what big companies pushed PHP, Ruby, Elixir, Python. Not to mentioned the continued use of C/C++ which persists specifically because it is ubiquitous. There are plenty of other smaller languages which you don’t hear about and aren’t backed by large companies but still do fine because they have a niche community.
In contrast big company languages often don’t get much traction because they are tied into the business plans of their masters or are pushed for reasons besides technical merit. Dart is a recent example of this.

>reference counting
>good

If using Swift guarantees the death of Rust and the suicide of all trannies everywhere, I'll make the switch

No, you waste of trips, it guarantees all those people coming to Swift.

Attached: 2013-05-23-582-Invaders-must-die.png (768x777, 903K)

Trannies aren't mentally stable enough to switch a language though.

>mentally stable
>to switch languages
pic related

Attached: 2016-01-26-809-Hoe-vs-Spade.png (1020x374, 471K)

>user who proved it in this thread
It is not possible to prove what is inherently an opinion. That Tim cock seems to have damaged your brain cells.

>t’s a vestige of an old age, like base 0 indexing.
swift users are genuinely retarded

This language indeed has one or two features/design that are interesting on the OO scope, but it fails horribly on everything else because it wants to try implement all the programming concepts available in the market.

>year 1967 +52
>fast, simple cycle detection algorithms for over half a century
>unironically defending Apple, one of the biggest companies in world, not having cycle detection.
Up to this point, I was considering giving swift a try. You have clearly blinded yourself to reason that you cannot even criticize swift.
You understand that you have adopted a cult mentality right? Steering away from this sinking ship. I suspect you are representative of a large portion of the swift community.

>vestige of an old age, like base 0 indexing
Ask me how I know you're a mathlet.

next fast.ai course will have 2 lessons on swift
I am learning this shit now

God why can't we just leave stupid fucks like you in the middle ages. Or even better, just ship all of the Jow Forumstards to india so they can be stupid with the rest of the world's stupid people

First I hate Apple so don't confuse it with Swift. I also hate Google but like using tensorflow.
As you can see here what you said isn't true there's cycle detection.
There are some things to criticize swift for.
I unironically think that before it reaches stable ABI on linux it's a piece of useless shit for anything interesting(and this has not happend yet).
But these posts are just untrue bs. It's like criticizing that C++ doesn't have pointers while it has them.
I am also not swift programmer and I am not interested in iOS/mac dev.
But using swift for ml servers looks super comfy.

PFFF.
Swift wastes so much time workaround around the problems of nullability that a well designed language wouldn't even have in the first place.

>open source
aka just shitting out source code in zip files
>cross platform
what a joke. it is works fine on mac but it is buggy as hell for linux and doesn't even work on windows.
>syntax isn't shit
t. ++i++ is too hard and C-like
>literally llvm lang with proper optimisations
literally slower than java lmao
>google choose it as the main tensor flow language
source on this?
>the creator is literally Clang creator and thinks X is going to be meme
nothing of value was added, literally
>the creator works in google
nothing of value was added
>Already works on Apple platforms
To fix it for you: only works on Apple platforms.
>IBM and hurr durr think
it is literally nothing
>botnet made it's fork for swift if other botnet does something unagreed upon
>it can use all of python libraries
are you implying other langs can't? you can even execute js in C++ and use any dynamic lang with C#'s dynamic type
>the biggest meme competitor bad we good
fuck off retard

This. People always talk about how "swiftroutines" make concurrency easy but if you can't solve the fundamentals in a single thread how do you expect to solve them by going to a different thread? Language designers need to solve "if err != nil" before they can solve thread starvation.

It works on windows though.
I have used vapor for one of my servers and everything went smooth(not sure how is Kitura, Perfect).
Rest of your points was already discussed in thread read it before posting.

they are good at other stuff and don't care about tech

>indexing arrays at 0 is stupid
Statement completely disregarded because of that. You're retarded

The more math you know, the more you come to hate adding or subtracting 1 from everything to make your equations work.

They are because, they still using base 0 indexing without any pointer artithmatic to justify it.

media.ccc.de/v/35c3-9670-safe_and_secure_drivers_in_high-level_languages

Our drivers in Rust, C#, go, and Swift are completely finished, tuned for performance, evaluated, and benchmarked. And all of them except for Swift are about 80-90% as fast as our user space C driver and 6-10 times faster than the kernel C driver.

news.ycombinator.com/item?id=19519164

Base 0 triggers the cnile. They can’t recognize that base 0 indexing lost its justification after we stopped requiring pointer arithmetic. In our programs.
Most important algorithms in maths and statistics start with 1 and not 0. Adapting them in languages in base 0 indexing is a common cause of errors when someone forgets to add or subtract 1 from an index, and that’s boilerplate which could be easily avoided.

You're wrong.

Got any compelling reasons for that or does starting at 1, like normies do, just make you uncomfortable?

not the guy you replied to but yes
cs.utexas.edu/users/EWD/transcriptions/EWD08xx/EWD831.html

The curios thing about that paper is Dijkstra recommending 0 -> N to avoid the error of using the wrong upper bound in a loop and improve clarity. The irony being that we then need to refer to the first element existing at the 0th index instead of the 1st as would be natural. And we still run into the bound error for series which start with 1, requiring us to either ignore the first element in an array, which often leads to bugs, or offset everything by 1 which also leads to bugs.
A sensible compromise would be to allow arbitrary indexing when defining an arrays in high level languages so that the indexing could match the use case. But few languages allow that flexibility.

wew lad C# is actually fast and go's latency is amazing

does this guy have a phobia against noses

He's afraid that if he draws them he'll wind up on Tumblr

confusing lol. i'm not even a programmer and i know what it does.

does the language come with balding trannies?

Swift is by far superior to C# and Java for mobile development, imbecile.
ARC dominates tracing GC on ARM, and ARM is the future of consumer electronics.

>mobile development
Kill yourself subhuman.

>Rust(great language btw.)
opinion_discarded.jpg

>not making mad dosh on apps
lmaoing at you poorfag

>making money on apps
Only Indians can do that anymore

You can't make money on the app store anymore.
The difference between the cost of 1 cent and (apparant) 0 is astronimically much for the normie.

The only way to make money with apps nowadays is to have something else (preferably a $9.99 subscription) to sell via that app. And to somehow try to sell it without using the IAP APIs so you won't have to give 30% to the people hosting the app store and to employ legal people who try their hardest to make sure your app won't be taken off the store for violating the "You must use In App purchases for every transaction and you may not link or mention any promotion in your app if you don't"

What else is there? Having a price tag is not viable, having a free-premium model sounds good until you have to decide - will I have a free version so good that people won't buy the premium version or will the free version lack a certain, important feature and you will get shit on with 1 stars from Indians and Pakis complaining that $1.49 is "way too expensive" for the app you spent months of your life on.

Hoping for donations from people whom your app helped is going to make the same amount as releasing the app for free.

So tell me. How do you make money writing apps? Especially now that every utility imaginable is on every app store (Except apple's, who deny useful utilities and make many useful utilities impossible to write thanks to iOS' primitive backgrounding and task scheduling)?